Introduction: Artificial intelligence (AI) has demonstrated its capacity and importance today, through its presence in multiple areas of contemporary society. Especially when talking about digital aspects such as threat detection in cybersecurity, where it offers many advanced tools to face modern cyber-attacks. Methods: This research is based on a Systematic Literature Review (SLR), taking into consideration the phases of Barbara Kitchenham's methodology, on the impact that AI has on cybersecurity, which seeks to identify the most effective techniques, their most important advantages and limitations, and the risks associated with their use. Key research questions include identifying techniques for early threat detection, comparing their effectiveness against traditional methods, and the risks inherent in their implementation. Results: In a digital environment where threats evolve rapidly, this type of intelligence allows organizations to identify and mitigate vulnerabilities more effectively than traditional methods, which often rely on static rules and signature detection. Although its implementation in the prevention of cyberattacks presents certain limitations, the effectiveness of intelligent systems depends largely on the quality of the data used for their training. Conclusions: As a main conclusion, AI represents a powerful technology in the fight against cyber threats, whose integration must be managed to maximize its effectiveness and minimize its defects. However, the deployment of AI in cybersecurity is not free of limitations such as the existence of vulnerabilities that invite hostile cyberattacks aimed at hijacking the operation of systems that use this intelligence.